[ https://issues.apache.org/jira/browse/SOLR-7142?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Shalin Shekhar Mangar reassigned SOLR-7142: ------------------------------------------- Assignee: Shalin Shekhar Mangar > TestFaceting.testFacets senseless > ---------------------------------- > > Key: SOLR-7142 > URL: https://issues.apache.org/jira/browse/SOLR-7142 > Project: Solr > Issue Type: Test > Components: faceting > Reporter: Michal Kroliczek > Assignee: Shalin Shekhar Mangar > Priority: Trivial > Attachments: testfaceting.patch > > > Hi, I am new to Solr sources and I was digging in facet tests. > I found unreasonable block of code in TestFaceting.testFacets: > {code} > ... ( 1st document is added and assertion is done) > // test gaps that take more than one byte > sb = new StringBuilder(); > sb.append(t(0)).append(' '); > sb.append(t(150)).append(' '); > sb.append(t(301)).append(' '); > sb.append(t(453)).append(' '); > sb.append(t(606)).append(' '); > sb.append(t(1000)).append(' '); > sb.append(t(2010)).append(' '); > sb.append(t(3050)).append(' '); > sb.append(t(4999)).append(' '); > assertU(adoc("id", "2", "many_ws", sb.toString())); > assertQ("check many tokens", > req(**"q", "id:1"**,"indent","true" > ,"facet", "true", "facet.method","fc" > ,"facet.field", "many_ws" > ,"facet.limit", "-1" > ) > ,"*[count(//lst[@name='many_ws']/int)=5000]" > ,"//lst[@name='many_ws']/int[@name='" + t(0) + "'][.='1']" > ,"//lst[@name='many_ws']/int[@name='" + t(150) + "'][.='1']" > ,"//lst[@name='many_ws']/int[@name='" + t(301) + "'][.='1']" > ,"//lst[@name='many_ws']/int[@name='" + t(453) + "'][.='1']" > ,"//lst[@name='many_ws']/int[@name='" + t(606) + "'][.='1']" > ,"//lst[@name='many_ws']/int[@name='" + t(1000) + "'][.='1']" > ,"//lst[@name='many_ws']/int[@name='" + t(2010) + "'][.='1']" > ,"//lst[@name='many_ws']/int[@name='" + t(3055) + "'][.='1']" > ,"//lst[@name='many_ws']/int[@name='" + t(4999) + "'][.='1']" > ); > {code} > This code adds second document, never commits and executes second query with > q=id:1 again. > So test will pass even if: > 1) Change assertions to ID's not related to 2nd document (like 2001) > 2) Add commit after 2nd document added (still filter is set to id:1) > 3) Remove at all creating 2nd document. > Patch changes test flow to fallowing: > 1) add document1 and commit > 2) assert facets to query *:* > 3) add document2 and commit > 5) assert query *:* to have facetes with values 1 and 2 when expected -- This message was sent by Atlassian JIRA (v6.3.4#6332) --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org